S. Dhinakaran

Influence of Buoyancy on Vortex Shedding and Heat Transfer from a Square Cylinder in Proximity to a Wall

The flow and heat transfer past a heated cylinder of square cross section mounted horizontally above a plane wall and subjected to a uniform shear flow is considered. The flow field is considered for a moderate range of Reynolds number (based on the incident stream at the cylinder upstream face and the height of the cylinder) and Grashof number at cylinder-to-wall gap height 0.5 times the cylinder height. The flow field and heat transfer are computed through a pressure-correction-based iterative algorithm with the QUICK scheme in convective terms. The code is tested for accuracy by comparing with published results for certain values of the flow parameters. We examine the combined effects of buoyancy and shear flow on the vortex shedding behind the cylinder in close proximity to a plane wall. A boundary layer develops along the wall, and this interacts with the shear layer formed along the two sides of the cylinder. The role of the thermally induced baroclinic vorticity production term on the vortex shedding is examined. Our results show that the breakdown of vortex shedding takes place beyond certain values of Richardson number (which depends on the Reynolds number). The dependence of the average rate of heat transfer from the surface of the cylinder on Reynolds number and Grashof number is also investigated.

Discrete Phase Modeling of Nanofluid Flow around a Circular Bluff Body

Extended Abstract Forced convection heat transfer around a circular bluff body is a scenario that occurs in many engineering applications such as electronic cooling, nuclear reactors, heat exchangers and so on. This classic area of research has gained much more momentum with the advent of nanofluids which are promising alternatives to conventional coolants due to their superior thermal properties. Many numerical studies on nanofluid flow and heat transfer using Single-Phase Modeling (SPM) and Multi-Phase Modeling (MPM) with Eulerian-Eulerian approach are available in literature. However, numerical studies using Discrete Phase Modeling (DPM) with Eulerian-Lagrangian approach which has the advantage of tracking individual particles and accounting for the interactions and momentum transfer between the fluid and particles are very limited [1]. In this study, a steady, laminar and forced convective nanofluid flow around a 2-D circular cylinder has been numerically analyzed using Discrete Phase Modeling (DPM). Steady state governing equations of flow and heat transfer for the basefluid and the equation of nanoparticle motion were solved using a Finite Volume Method (FVM) and trajectory analysis approach, respectively. An Eulerian approach was employed for the basefluid, while a Lagrangian approach was used for calculating the particle trajectory with two-way coupling. The continuous phase (basefluid – H2O) was considered to carry the discrete phase (nanoparticle – Al2O3) and the momentum transfer between the nanoparticles and basefluid was also taken into account. The nanoparticles were assumed to be randomly dispersed in the continuous phase and the influences of Brownian motion, thermophoresis, drag and nanoparticle weight were also considered. The flow and heat transfer was analyzed at 10 ≤ Re ≤ 40 and nanoparticle volume fraction varying from 0% to 5%. In general, the heat transfer increased with the addition of nanoparticles and the augmentation in heat transfer was more pronounced at higher Reynolds numbers. The influence of reflect and trap boundary conditions over the cylinder was also analyzed. The results of DPM were compared with that of the SPM approach. The results of DPM showed that the nanoparticle concentration is not uniform, which is contradictory with the basic assumption of SPM. In fact, a drop in concentration near the walls was observed for trap boundary condition whereas, a slight increase in concentration was noted in reflect boundary condition. Furthermore, addition of nanoparticles resulted in a marginal increase in total drag coefficient. Brownian motion exhibited more influence on the heat transfer augmentation than thermophoresis. It is believed that the Discrete Phase Modeling (DPM) using Eulerian-Lagrangian approach with two-way coupling is more realistic than other two approaches. But, it is also necessary to mention that the SPM and MPM are simpler and computationally less expensive than DPM.